Sie sind auf Seite 1von 3

SWARM INTELLIGENCE

K. Harish

Department of Electronics and Computer Engineering


Sreenidhi Institute of Science and Technology
Yamnampet,Ghatkesar Mandal,Medchal Dist,501301,Telangana,India
harishkurakula18@gmail.com

1. Abstract: last model Kennedy’s Adaptive culture model which tells that
there is no effect of similarity on interaction.
The emergent collective intelligence of groups of simple
agents. A population of interacting that optimizes a function Then comes the ant colony optimization by Dorigo in 1992
or a goal by collectively adapting to the local and/or global and particle swarm optimization by Kennedy, Eberhart & Shi
environment. Swarm intelligence is not about how collections in 1995.
of cells yield brains, and it is not about how individuals form
civilizations, and it is not about such things as the lifecycle of
the slime mould. However, it is about individuals cooperating 5. PRINCIPLES OF SWARM INTELLIGENCE:
to achieve a definite goal. Such as, ants finding the shortest Swarm Intelligence is mainly implemented by the basic
path between their nest and a good source of food, or bees principles and these principles are called as the principles of
finding the best sources of nectar within the range of their Swarm Intelligence. The first principle is the Proximity
hive. These similar natural processes have led directly to Principle which means that the population should be able to
families of algorithms that have proved to be very substantial carry out simple space and time computations. The next
contributions to the sciences of computational optimisation principle is the Quality Principle which means that the
and machine learning. population should be able to respond to quality factors in the
environment. The next principle is the Stability Principle
which means that the population should not change its mode
2. INTRODUCTION: of behaviour every time the environment changes and the last
Swarm Intelligence is the collective behaviour of principle is the Adaptability Principle which means that the
decentralized, self-organized systems, natural or artificial. population must be able to change behaviour mode when its
Swarm means to move somewhere in large numbers. It is an worth the computational price.
innovative distributed intelligent paradigm for solving
optimization problems. It is taken from the biological By using these above principles they introduced the two
examples by swarming, flocking, and herding phenomena in optimizations called the Ant colony Optimization and the
vertebrates. Particle Swarm Optimization.

A. Ant Colony Optimization


3. PHILOSOPHY: Ant colony optimization is introduced by Dorigo in his
The concept is employed in work on Artificial Intelligence. doctoral dissertation, is a class of optimization algorithms
The expression was introduced by Gerardo Beni and Jing modelled on the actions of an ant colony. It is a probabilistic
Wang in 1989, in the context of cellular robotic systems. They technique useful in problems that deal with finding better
consist typically of a population of simple agents or bodies paths through graphs. Artificial ants are treated as simulation
interacting locally with one another with their environment. agents who locate optimal solutions by moving through a
Interactions between such agents lead to the emergence of parameter space representing all possible solutions. Natural
“intelligent” global behaviour. ants lay down pheromones directing each other to resources
while exploring their environment. The simulated ants
4. HISTORY: similarly record their positions and the quality of their
solutions, so that in later simulation iterations more ants locate
Latane’s dynamic social impact theory was that the for better solutions. This can be mainly implemented in
Behaviours of individuals can be explained in terms of the telecommunication networks which are also called ant-based
self-organizing properties of their social system and the next routing.
comes the Axelrod’s culture model which tells that the
population of individuals are pictured as strings of symbols, or This is pioneered separately by Hewlett Packard in the mid
“features” and the probability of interaction between two 1990s. This ant based routing uses a routing table reinforcing
individuals is a function of their similarity and then comes the the route successfully traversed by each ant which floods the
network. As the system behaves stochastically and is therefore Fig. 3 The equation required to calculate the fitness of a particle.
lacking repeatability, there are large hurdles to commercial
deployment.

Fig. 4 The Swarm Technique.


Fig. 1 A natural behaviour of ant which lead to an idea of ant colony
optimization. 6. APPLICATIONS:

There are various applications on this technique. Mainly the


U.S. military is investigating swarm techniques for controlling
unmanned vehicles. The European space agency is thinking
about an orbital swarm for self assembly and interferometry.
NASA is investigating the use of swarm technology for
planetary mapping. This technique has also become the
possibility to control nanobots within the body for the purpose
of killing cancer tumors. This has also been applied for data
mining. Other applications are - Crowd Simulation, Human
Swarming, Swarm Grammar, and Swarmic Art.
Fig. 2 Ant Colony Optimization
7. CONCLUSION:
B. Particle Colony Optimization Swarm intelligence is the discipline that deals with natural and
Particle Swarm Optimization is a computational method artificial systems composed of many individuals that
that optimizes a problem by iteratively trying to improve coordinate their activities using decentralized control and self-
a candidate solution (inequalities, equalities and integer organization.
constraints) with regard to a given measure of quality. It is
developed by Kennedy, Eberhart & Shi in 1995. It is related to In particular, the discipline focuses on the behaviour of social
bird flocking, fish schooling. This process is taken by the insects such as fish schooling and bird flocks and colonies of
following observations, first we need to initiate the population ants, termites, bees, and wasps. The most well-known
in hyperspace and then evaluate the fitness of individual example of system studied by swarm intelligence is particle
particles. If there is a mismatch in the velocities, then modify swarm optimization (PSO) and the Ant Colony Optimization
the velocities based on previous teat and global or (ANO).
neighbourhood best. After this terminate this on some
condition and then repeat the above steps for the next particle. 8. REFERENCES:
[1]. https://en.wikipedia.org/wiki/Swarm_intelligence#A
pplications
[2]. Swarm Intelligence book by Kennedy and Eberhart
[3]. Swarm Intelligence for Analyzing Opinions in
Online Communities
[4]. Intelligence with Spiking Neural Achieving Swarm
Oscillators
[5]. “Implementation of Evolutionary Fuzzy Systems”
Authors: Shi, Eberhart, Chen IEEE Transactions on
Fuzzy Systems April 1999
[6]. World's Largest Swarm of Miniature Robot
Submarines

10. AUTHOR PROFILE:

K. Harish – I am an engineering student of second year


studying in Sreenidhi institute of science and technology. I am
working as a bug reporter for miforums for the year 2018.

11. PLAGIARISM:

This project was written by me and in my own words, except


for figures from published and unpublished sources.
Plagiarism ratio = 70:30

Das könnte Ihnen auch gefallen